The Business of Credit: 101 for Startups

Yash Jain
Published in
14 min readNov 26, 2020


There are more than 100+ startups doing lending in India and then there are many established ones who are thinking of building credit as an adjacent business model for improving top line or bottom line metrics or may be just for better customer stickiness. There is a lot of buzz these days about credit. And If you are one of those startup executives or founders, you must be asking yourself why should I even bother reading or even skimming this yet-another-article from some-random-guy?

I have deeply looked at more than 20 credit startups over the last 5 years and have closely worked with one of them (a portfolio company) in raising $100M+ in equity and debt. Have also closely studied listed Chinese credit startups such as Lexin Fintech and Qudian.

Enough said, you may want to give this a little more attention. This is a rather longish read. I am going to talk about the key concepts of building a credit business — not chasing vanity metrics, building debt, business model, unit economics, and other things but not limited to valuation.

Note that this post is about Balance Sheet run credit businesses and not platform/technology or infra plays.

Don’t mistake demand for growth

Credit is a pull product. There are enough people out there who are always looking for some money to do something or spend somewhere. The question is whether you want to lend them or not; what is the level of risk you are willing to take.

Credit companies use credit scores as a way to throttle this risk. They call this underwriting. Having said that, India`s most prized consumer lender Bajaj Finance lends primarily to prime customers (CIBIL 700+). And, it continues to grow that business at a humongous pace courtesy to the partnerships, service, and the brand it built over years. People like you and I, who are reading this article, want to go to Bajaj when we are in need of some credit buying our next iPhone.

Bajaj may chose to lend to a newer class of more risky customers, by tweaking their risk model and factoring that in their pricing accordingly (we will talk about this later). Voila! People will come running; their loan book will increase. But I hope, they can get their money back.

Yes, that is precisely my point.

Lending money is one thing, getting back is another. When startups get excited about their ability to disburse money at high monthly growth rates, they should wait a while to see if they can get the money back at industry rates or even better: complete the money cycle.

All the more, investors, especially debt ones, are looking for startups` ability to repeatedly deliver such successive money cycles. You could disburse $1B in a month if you wanted to, by throttling your risk profile; it is not enough. People want money, only that traditional established lenders are not giving them any because of their credit profile. Thus, in my view, having figured out a way to disburse capital is not really a moat in this business.

This also means that, by design, you CAC should be low. If you are spending a lot of money convincing people to take money from you, there is something wrong; I have seen businesses where CACs are at ~$75 for a ~$100 loan. Not that exciting.

“Our disbursements are growing at 50% month over month” isn`t a exciting statement anymore and the industry should force people to move away from it. Maybe the better statement is “Our money cycles are growing at 50% every 3 months” implying that you can collect money as well.

Credit is a supply side (debt) business

A great credit company is built on the back of quality of debt. You cannot fuel this business on equity alone. Bajaj Finance has ~$15B in debt compared with $4.5B in equity on its balance sheet; $3.5 in debt for every $1 in equity — one of the best in the industry.

It is safe to say that founders and management teams get the equity game; a lot has been written or said about that. Let me use this section to talk about the other side of this coin — debt.

You have just designed your credit product (some ideas here), and you have been doing some lending through the equity you raised from a VC. You start approaching debt investors for credit lines. The cost of debt that you will be able to tap into will be very high — more than 20% per annum.

Why? Because you are getting credit after multiple hops. Let me then come to one of the most important topic of this post — credit hop.

The largest debt pools or players in the market are banks. They take deposits and lend to large corporate, small businesses, and consumers, keeping a mix that suits their risk profile and business strategy. Now, Indian banks have traditionally not lent (not in large volumes) to small businesses and consumers because of many reasons — lack of data, cost of onboarding and servicing the customer at lower price points, and ability to collect disburse money effectively.

So where do these people or businesses get money? Many of them don`t and many of them borrow from informal sources at obscenely high rates. A world bank report estimates that 45% of adults in India have borrowed while only 7% of adults have done so from formal institutions. For this small proportion, the credit journey isn`t that straightforward. Talking about that would be digressing.

Capital hops in a typical consumer credit set up

For those who do get money from formal channel, it comes at a higher cost as result of credit hop; Banks lend to NBFCs and NBFCs thus lend to these riskier customers, pricing high for associated risk. Banks are happy taking an exposure to that market via an indirect approach. Your startup will probably come after NBFCs, adding an additional hop to this credit chain. A typical credit hop in an unsecured consumer lending startup may look like the one in the image to your left. So that`s that, you are not getting your first debt investors cheap. Note: NIM stands for Net Interest Margin — what you lend at minus what you borrow at, thats your spread.

Business model: Start high, aim lower

Let`s shoot straight to the point here.

Start with pricing high and lending to riskier customers — new to credit & sub prime. Subsequently, build your moat around underwriting capability and keeping the costs very low by running a lean shop. Overtime, get your cost of capital lowered (move up in the credit hop). As a result start lending to near prime and prime customers, letting your brand, and digital convenience & experience, speak to the customers. This approach is in fact executed well by a few Indian startups and consequently they have done well. This journey could (and should) look like this:

As you can see, this hypothetical startup is able to work on its cost of capital — by convincing debt investors of the money cycles, but only slowly. This makes me come to my next point.

Debt investors worry about getting their capital back in next 6/12/18 months.

Investors in Bombay think differently from those in Bangalore. Consequently, your debt profile will grow much slower than the equity profile, making your leverage ratio (debt/equity) only creep up slowly. Remember, Bajaj is a ~3.5x. “Good, you returned our $1M principal with interest; our investment committee has decided to increase our exposure to your company to $2M” (and not $20M); they usually operate like this. “Lets see if they can maintain similar collection rates or money cycles at twice the scale,” they would often think.

Thus, a good lending startup will see something like this happen on its Balance Sheet. Credit is a very much balance sheet business as it is a P&L business. The knowledge of balance sheets is underrated in our startup ecosystem which comes haunting when you run a credit startup.

Your ideal path for capital would be through a Tier 2 NBFC, Tier 1 NBFC, Banks, and then deposits.

Please note that steering through this process, you will have to undergo hordes of regulatory and compliance milestones.

  1. Getting your NBFC license.
  2. Becoming a systematically important NBFC (NBFC S.I.) when your turnover exceeds 500 Crore: more scrutiny your way. Large debt investors rely on such categorisation to invest in you. At this point, expect regular meetings with local RBI officers.
  3. Getting yourself a credit rating. This needs history of loan cycles and rating guys are even more ruthless. Banks typically do not invest in companies without a credit rating in place.
  4. Becoming a deposit taking NBFC (NBFC-D) like Bajaj Finance. This again requires lot of history, profitable P&Ls, de-stressed balance sheets, and some bit of lobbying.
  5. Maybe, getting a banking license (yes, this would be dream come true). Banking licenses dont come easy in India and they bring with them lot of regulatory oversight and commitments towards inclusion — opening up rural branches, ATM network etc.

I would also like to make a point about digital convenience. Many startups make the mistake of counting digital convenience as their sole right to win. Well, as you can see from a lot of what I have talked about, digital convenience comes when you are competing with prime customers and you are not going to get there without traversing the debt cycle.

You will have to build a fundamentally sound, positive unit economics business right from Day-1.

Otherwise debt investors are not interested. And you cannot build this business alone on equity unless, of course you have SoftBank by your side, which I think isn't a luxury anymore.

There are some startups which focus on secured asset classes such as Gold Loans and Home Loans in which they can start a little higher in the credit hop chain. Having collateral (gold or real estate) provides some form of risk mitigation but the basic principles of building money cycle history and consequently your debt profile remain the same. Of course, this comes with an added complexity of managing the asset — valuation, holding, sale in case of defaults — for the customer.

How to hack this business model — shorter tenors, low operating costs, and experienced team.

As mentioned above, building your debt profile is a slow process. Let us say your product is a 4-year education loan. It will take 4 years for you to complete a money cycle and convince your investors to increase exposure. Of course, you can show partial collection in form of instalments to make a case, but it is still slower. There is no harm in that but then you have to get your expectations right — you ain`t going to be a high growth, headline making credit startup.

Chinese credit startups that have done well started with 7-day, 15-day loan products, iterated fast, and thus, convinced debt investors of their money cycles` robustness. As their capital costs went down, they started offering longer tenor products, lowering their pricing and keeping regulators happy. We will talk about this in the unit economics section, you will see that shorter tenors have a bearing on unit economics in a big way.

Next, keep your operating costs extremely low. Let us talk about costs first — The 4Cs. You will have cost of acquisition, cost of capital, cost of credit (NPAs),and cost of operations.

Like I have said before, as a pull product your CAC or cost of acquisition should not be unusually high. If it actually is, make sure you have the right marketing team which understands doing this better. We have talked about cost of capital in great deal above.

Cost of credit or delinquency cost is about the quality of your underwriting and collection. As a tech company, you will (and should) be able to build a great underwriting model using conventional and alternative sources of data. This one is really your mojo. What could blow up is your collection operations — you don`t collect as much as you should and you spend a lot of money collecting what you do — which effects your cost of operations where payroll of collection team and/or commission of collection agency sit(s). Thus cost of credit also has an impact on cost of operations which we will come to now.

The cost of operations is where your digital-credit shop has the biggest operating leverage. While traditional lenders use manpower to analyse credit application, appraise existing ones, often taking days, you could have a completely digital workflow run by a few people. The costs you will incur shall include cost of data sources, usually — CIBIL reports, PAN and Aadhar data sets; and cost of digital KYCs — video verification etc. Most of these services are provided for by many startups out there on fully variable basis (APIs). The big manpower would go into collection operations which we have talked about before. You will have to spend your energy on making sure that majority of your collections are soft in nature (phone calls, notifications, emails, often all automated). The moment you get into hard collection bucket — home visits, legal routes etc, your costs will shoot up. Note, your quality of underwriting will keep the need for excessive hard collection low. So, collections and delinquencies go hand in hand.

Last, hire an experienced team. While young founders like to believe that they can show up in their casuals and get investors excited with their sheer wit, passion, and ability to figure things out. This does not work when you have to deal with debt investors. They are in business of getting their money back at standard returns. They are not looking for outliers but averages. Thus, they like to trust (and bet on) people who have had relevant experience. Having said that, if you are a young founding team, your CFO becomes the most crucial hire; make sure you get somebody experienced.

Unit economics should bake in multiple loan and tenor expansion

I know you must be tired reading this long post. But, wait, I have come to the most exciting part now — unit economics. I have already explained key cost elements above. With that pretext, I have taken the liberty of drawing a sample unit economics table for a consumer lending startup below.

As you can see, this company is acquiring risky customers, taking a risk on them, and taking a percentage of them who don`t default towards the second loan, so on and so forth. It is this cohort that will make your business sustainable — do you have a significant percentage of customers in your cohort who go to fourth loan (in the above case) to make your business model viable?

In the above case, by the fourth loan, you make a $7 in Life Time Value (LTV) at $1 CAC. This means that if 1 out of every 7 customers acquired, graduates to the 4th loan, you should breakeven. The more the better.

What you must also note is that, as this hypothetical company is taking these customers forward in the credit lifecycle, both capital exposure in terms of amount lent and tenor of the loan are increasing, taking your IRRs down. For example, 5% processing fee for a 30-day loan means an annualised 60% IRR, add to that 25% yield/price, and your IRR is 85% which regulators dont like. Credit Card IRRs in India range between 45–50%

Do this (the 4–5 loan cohort cycle) for multiple cycles of money, and your debt investors will agree to give you the next tranche at lower cost of capital, and you get on this coveted upward spiral of credit hop.

Okay this is my last topic of the post: The valuation faux pas.

Well, you are used to seeing high growth startups raise money at crazy valuations; often at very high revenue multiples and even crazier EBITDA/Gross Profit multiples. Sure, you will see that in credit business but only for few rounds — maybe up to Series B or C. Your early rounds mainly are a function of incoming investor wanting to hold a certain ownership — 15%, 25%, 30%, 33% etc.

Growth equity investors start taking multiples seriously, especially in credit businesses. Why so serious, you may ask? Because they know that someone who will buy or fund this company down the line will.

Lending multiples are ruthlessly based on the health of your P&L and Balance Sheet — on your Net Worth (Assets — Liabilities on your Balance Sheet). Typically companies are funded at 2–4x of Net Worth in the Indian market, depending on your margins, cost of credit, risk profile etc.

I have also seen people valuing lending companies at some 10–15x disbursements. To me, that is just bizarre because for the longest time I have tried very hard to understand what does disbursements have to do with Net Worth or the margin structure of the business. Again, in my humble opinion, they are doing the mistake of valuing companies on vanity metrics — GMV equivalent Disbursement. And as I have said before, let us not mistake demand for growth; disbursements along don`t mean a thing!

Okay, so let us take an example. Suppose you are this high profile team which has raised $50M in three rounds and your last round was at $150M. Sure, you will have the pressure to continue raising money at high valuations in future. But there will come a time, where some downstream investor will value you at 3x Net Worth. So to be a $1B company, you will need to have $330M in Net Worth, that means you will have to generate $330M in profits to be put in reserves and surplus of Balance Sheet (assuming your previous capital was spent just funding CAC and didn't generate substantial cash) outside of investing some cash back into your business. Well, that sounds daunting isn`t it?

The question, all of us, investors and founders alike, should ask ourselves that how will these private market valuations based on vanity metrics converge with more saner and realistic valuations based on real P&L and BS metrics. How do we avoid this faux pas?

That is why I stress on building a fundamentally profitable business from Day-1 when your choice of sector is credit.

You could argue on the other hand that you are a platform and you can do this and that with the data you will gather. Sure, make that choice early! But if you are going to be a business where majority of revenue comes from borrowing money and lending it to people, making a spread in between, be prepared to be realistic about valuations.

It is yet to be seen, where these credit businesses, which position themselves as platform with some kind of Facebook like network effects, will go. Because there isn`t enough data out there to base any kind of assumption. Just because some investor across the continent valued a company using some multiple should not be our answer to valuations. We dont actually know what was the thesis there.

I know I have said a lot, and there is a lot of digest in one post. So, if you wanna talk about something specific, do reach out. I am always up for challenging my assumptions and learning more myself.

And if you are a founder who enjoys working with thesis driven investors, do reach out to me at I run a fund called Sparrow Capital where we invest in early stage Fintech, SaaS, and B2B companies.

Talk to you soon!



Yash Jain

Founder and Managing Partner at Sparrow Capital. Excited about the role of technology in changing lives of people, for better.